методология / haritonchik / харитончик книга по словообразованию

Elisabeth O. Selkirk*

A GENERAL THEORY OF WORD STRUCTURE

Within the context of generative grammar, a variety of approaches to morphology have been pursued. In the first work on the topic, Lees proposed that complex words – compound words as well as those involving derivational or inflectional affixation – be derived through the operation of syntactic transformations from deep structures including only noncomplex words. Chomsky (1970) presented important arguments against this approach to derivational morphology, concluding that derivationally complex words must be present in deep structure. The same sorts of arguments lead one to conclude that compounds are present in deep structure (cf. Allen (1978)). In this monograph, as in other generative works on morphology, the conclusion that words with derivational morphology and compound words are not formed by syntactic transformation is taken as a point of departure. Along with this view, I adopt the somewhat less universally held assumption that inflectional affixation is not accomplished by syntactic transformation, but that, with derivational affixation and compounding, it instead forms part of a morphological component of grammar.

My purpose is to examine what I will call the syntax of words, by which I mean the structure of words and the system of rules for generating that structure. While much has been said in the recent linguistic tradition about the syntactic structures of which words form the basic units, considerably less attention has been paid to the structure of the words themselves. Perhaps this has come about because word structure seemed perfectly obvious, apparently a mere extension of syntactic structure. This is the view of word structure implicit in Chomsky and Halle (1968; hereafter SPE), for example, and one that is adopted in most subsequent works.

However, it is an error to view word structure as merely the “lower” portion of a syntactic representation that is entirely homogeneous in character. It can be argued

* Selkirk E.O. A General Theory of Word Structure // The Syntax of Words. The Massachusetts Institute of Technology, 1982. P.1–18.

160

that, aside from the category Word itself, the categories involved in word structure are distinct from those of syntactic structure and, moreover, that the two types of structure combine these categories in significantly different ways. It in fact seems possible to construe word structure as an autonomous system. In my view, the category Word lies at the interface in syntactic representation of two varieties of structure, which must be defined by two discrete sets of principles in the grammar. Yet I will argue that word structure has the same general formal properties as syntactic structure and, moreover, that it is generated by the same sort of rule system. <…>

A context-free rewriting system by itself is capable of generating all| of the words of a language, but only at a certain cost. Members of a certain class of morphemes, the affixes, display idiosyncratic distributional properties. The suffix - ity, for example, attaches only to an adjective and with it forms a noun: obesity =

N[A[obese]A -ity]N. The suffix -ify, on the other hand, always constitutes a verb along with a sister adjective or noun: codify = V[N[code]N -ify]V, purify =

V[A[pure]A -ify]V. The most obvious way of capturing these distributional peculiarities within a context-free rewriting system is to introduce each affix directly by a rule specific to it, as in (1.1), which means, in effect, positing a

separate rule for every affix of the language:

(Such a rule system would also involve rules rewriting the preterminal categories

assign affixes a categorial status and to capture generalizations about possible word structures in a direct way. <…>

To keep the focus on the issues which I consider to be central here – the issues of the nature of word structure and the rule system for generating it – I will assume, along with the various other generative theories of morphology, that the

161

rules of word structure form part of what one may call the lexical component or simply the lexicon (understood in a broad sense). As it is viewed here and in most earlier theories, the lexical component contains a variety of subcomponents. First, it contains a list of freely occurring lexical items (which I will assume to be words, in English). We may call this the dictionary (or lexicon, in the restricted sense). Second, it contains a list of the bound morphemes of the language. This, together with the dictionary proper, I will call the extended dictionary. Third, the lexical component includes the set of rules characterizing the possible morphological structures of a language, the word structure rules of the present theory. The word structure rules, along with the structures they define, are the central concern of this monograph. Together with the extended dictionary, they form the core of the word structure component of the lexicon or, shall we say, the morphological base. <…>

One characteristic that distinguishes morphology from syntax, to be sure, is the fact that many of the entities defined as well formed by the rules of morphology are fixed expressions. Most words we speak and understand we have heard before, while sentences are for the most part novel to us. More precisely, what distinguishes words from sentences is that most words are in the dictionary.

There are a number of reasons for saying that the list of items called the dictionary forms part of the speaker‟s knowledge (or grammar) of a language.

First, speakers have intuitions about what is or is not an actual word of the language (as well as intuitions about what constitutes a possible word of the language). Second, and more important, the individual characteristics of words are not always predictable. The meaning of a simple word is totally unpredictable, and even in the case of complex words the meaning often cannot he predicted on the basis of its component parts. The conclusion is that a word and its (idiosyncratic) meaning must be paired in a list; that list is the dictionary. Phrases whose meaning is not compositional – that is, those phrases that are usually called idioms – will also have to be listed in the lexicon.

Consider now the fact that the multimorphemic words of this list must be said to have an internal structure. Speakers have intuitions about the structure of

162

existing words of their language. These intuitions are presumably based on their knowledge of the word structure rules of the language, and indeed it seems that in general the existing lexical items of a language (more exactly, the words of the lexicon) have structures generable by the morphological component of the language. But the word structure rules cannot be viewed as generating these words anew each time they are used, for this contradicts the notion that they are listed; no distinction would then be drawn between existing and possible (or newly generated) words, and no means would be available for representing their idiosyncratic, noncompositional features. In the case of existing lexical items, then, it would seem appropriate to view the word structure rules as redundancy rules or well-formedness conditions on lexical items. More generally, it seems possible to impose the following condition:

For every word of the language, there must exist a derivation via the word structure rules of the language.

This condition allows us to treat existing words and possible words in uniform fashion. If a word (existing or possible) is to be well formed, its structure must be among those generable by the word structure rules of the language.

COMPOUNDING

Compounds in English are a type of word structure made up of two constituents, each belonging to one of the categories Noun, Adjective, Verb, or Preposition. The compound itself may belong to the category Noun, Verb, or Adjective. My purpose here is not to provide a thoroughgoing description of compounding in English (which the reader can find in Marchand (1969), Adams (1973), and Jespersen (1954)). Rather, I will focus on what I consider to be the essential features of English compounds and their relevance to the theory of word structure outlined previously. <…>

The vast majority of English compound types are headed – specifically, right-headed – and the heads of these compounds display the syntactic and semantic characteristics that are expected of heads. There are some compound

163

types which are clearly not headed. <…>

The Structure of Compounds

As we see from examples (1)–(3), a compound noun may consist of a noun, adjective, preposition, or verb on the left and a noun on the right, a compound adjective may consist of a noun, adjective, or preposition followed by an adjective,

and a compound verb may consist of a preposition followed by a verb:

164

compounding, to distinguish it from the sort of compounding of (in part) Greek origin, which is common in specialized, sometimes learned terminology: telescope, metamorphosis, erythrocyte, kilometer. Such compounding forms a discrete system.

The paradigms of compound types given in (1)–(3) contain several gaps. Among the missing are compound verbs and adjectives whose left-hand member is a verb: v[V V]v, A[V A]A. These simply do not exist in English. By contrast, the compound noun type N[V N]N is attested, though rare, e.g., swearword, scrubwoman. Also missing from the paradigms (1)–(3) are the verb compound types V[N V]V and V[A N]V. Indeed, this arrangement of the facts implies that the only verb-on-the-right verb compound type of English is the one consisting of a preposition plus verb. Admittedly, one might contest the absence of the V[N V]V and V[A V]V types from (3), for the language does contain verbs which seem to display this structure:

(6) Verbs

N V A V

a.globe-trot sharpshoot stage-manage dry-clean air-condition

window-shop mass-produce

house break

Marchand argues, however, that all of these are back-formations, which he terms pseudocompound verbs, coined with reference to already existing nominal or adjectival compounds. The words of (6), for example, were coined on the basis of those in (7):

166

If indeed, as Marchand argues, all compound verbs like those in (6) presuppose the existence of noun and adjective compounds as lexical items, then such types are to be distinguished in terms of their derivation from the compounds of (1)–(3), whose existence does not presuppose the existence of compounds of other types. While concurring with Marchand‟s assessment of verbal compounding in English, Adams gives some examples that appear to suggest a (limited) direct formation (not backformation) of verb compound types, e.g., chain-smoke. However, it is not clear that the sporadic existence of such types, an innovation in English, yet reflects a change in the basic rules for verb compounds in English. Assuming the correctness of

Marchand‟s claim, I have excluded the verb compounds of (6) from the paradigms

(1)–(3). These paradigms thus represent the types of compounds that the system of word structure rules generates directly, without recourse to back-formation. Backformation, however it is to be conceptualized (and formalized), is taken here to be a qualitatively different sort of phenomenon, not part of the strictly generative system of the morphological component, which consists of the word structure rules of the language.

The fact that there exist systematic gaps in the paradigms of compound types in English is of some importance. Insofar as these gaps can be shown to be

167

particular to English and not to follow from universal principles, the grammar of English must encode them. This means in particular that the grammar of compounding must explicitly mention the combinatorial possibilities of categories within the compounds belonging to the different categories Noun, Adjective, and Verb. In other words, the rules of the system must be formulated in terms of specific syntactic category names. <…>

The case that the gaps in the English compound paradigms do not follow from universal principles is easily made, for the compound types missing in English do occur in other languages. For example, verb compounds consisting of two verb constituents seem commonplace; they are found in as widely disparate languages.<…> Verb compounds consisting of a head verb plus an “incorporated” noun, presumably not derived via back-formation, are also not uncommon; v[V N]v compounds of this sort occur in Vietnamese (Thompson (1965)) and Chinese (Newnham (1971)), and v[N V]v combinations are found in Iroquois. Shoshonean, and elsewhere (Sapir (1911)). <…>

In sum, we have seen that a grammar of a language must include some system of rules explicitly demarcating the range of possible compound types of the language (by mentioning specific categories).

COMPREHENSION CHECK

1.What is meant by syntax of words? Explain the notion. On what ground can the combinability of derivational bases and affixes be termed syntax?

2.What is the author‟s idea about differentiation between morphology and syntax based on?

3.What is the basis of speakers‟ intuitions about the structure and meaning of derived words of their language?

4.Under what main condition are new language units structured?

5.What are the main structural types of compounds in English?

168

Rochelle Lieber

MORPHOLOGY AND LEXICAL SEMANTICS

In his comprehensive descriptive work on English word formation, Hans Marchand expressed the following opinion about the meaning of derivational suffixes (1969, 215): “Unlike a free morpheme a suffix has no meaning in itself, it acquires meaning only in conjunction with the free morpheme which it transposes.” In context, what Marchand means does not seem nearly so radical. He goes on in the same passage to explain that derivational suffixes change either syntactic or semantic class, and his prime example is the suffix -er (1969, 215):

As a word class transposer, -er plays an important part in deverbal derivatives, while in denominal derivatives its role as a word class transposer is not important, since basis and derivative in the majority of cases belong to the same word class "substantive" ...; its role as a semantic transposer, however, is different in this case. Although most combinations denote a person, more specifically a male person (types potter, Londoner, banqueter, weekender), many other semantically unrelated senses are possible. Derivatives with -er may denote a banknote, bill

(fiver, tenner), a blow (backhander), a car, a bus (two-seater, two-decker), a collar

(eight-incher), a gun (six-pounder), a gust of wind (noser, souther), a lecture at a certain hour (niher "a class at nine o'clock"), a line of poetry (fourteener), a ship

(three-decker, freighter, …).

Marchand of course does not mean to say that -er actually means “car,” “bus,”

“banknote,” or “gust of wind” in these forms. Rather he suggests that the meaning of the affix is fluid enough to allow all of these meanings in combination with particular bases. But why should this be? What, if anything, does -er add to a base to give rise to these meanings?

This book is about the semantics of word formation. More specifically, it is about the meaning of morphemes and how they combine to form meanings of complex words, including derived words (writer, unionize), compounds (dog bed, truck driver), and words formed by conversion. To my knowledge there is no

Lieber Rochelle. Morphology and Lexical Semantics. Cambridge University Press, 2004. P.1-12. 169